Migratory strategies, fat loads and torpor use in migrating hummingbirds
Data files
Feb 19, 2026 version files 331.37 KB
Abstract
North American bee hummingbirds frequently enter torpor during autumn migrations to save energy. The frequency with which they use torpor is mediated by fat loads. To estimate frequencies and temporal patterns of torpor use, we measured lipid loads and feather hydrogen stable isotope ratios (δ2H) in 942 individuals of four species of migrating bee hummingbirds: Selasphorus rufus (Rufous Hummingbird), Selasphorus calliope (Calliope Hummingbird), Selasphorus platycercus (Broad-tailed Hummingbird), and Archilocus alexandri (Black-chinned Hummingbird), at a stopover site in Southwestern New Mexico. In a subsample of juvenile individuals of each species, we constructed functions relating lipid loads and feather δ2H values to torpor incidence, using respirometry. We used these functions, along with lipid load and δ2H data from the remaining captured juvenile birds to estimate their temporal patterns of torpor use. Lipids were significantly higher in adult S. rufus and S. platycercus than in juveniles but did not differ with age in S. calliope and A. alexandri. In S. rufus, lipid loads decreased with capture date, while in A. alexandri they increased. There was no effect of capture date on lipid loads in S. platycercus and S. calliope. In all species torpor incidence decreased with increasing fat loads in a predictable fashion. A large fraction (70%) of S. calliope and a smaller fraction (30%) of S. rufus individuals entered torpor, but these fractions remained constant over time. In contrast, the fraction of S. playtcercus individuals that entered torpor increased with capture date, whereas that of A. alexandri decreased. Our results revealed similarities, but also unexpected interspecific differences in fat loads and torpor incidence in migratory bee hummingbirds. Some of the correlates of these differences remain unexplained but represent an opportunity to better understand the causes and consequences of torpor-mediated migration.
Migratory strategies, fat loads and torpor use in migrating hummingbirds
Dataset DOI: 10.5061/dryad.dbrv15fdk
Description of the data and file structure
This supplemental data (collected in the field by the authors) was used to analyze energy management strategies in migrating Selasphorus rufus, Selasphorus calliope, Selasphorus platycercus, and Archilochus alexandri hummingbirds. We captured 942 birds and performed respirometry trials on 94 adult and juvenile individuals. We measured both body lipid and feather hydrogen stable isotope ratios on all birds. We used results from respirometry trials of 63 juvenile birds to predict torpor frequency as a function of body fat and latitudinal origin on the larger subset of 573 juvenile birds. Measured variables and units are as follows:
Files and variables
File: Supplemental_Material_S1_Fat_and_Metabolic_Parameters.csv
Description: Full dataset of Metabolic parameters used in lipid (fat), hydrogen stable isotope, and torpor analyses in four species of migrating Selasphorus rufus, Selasphorus calliope, Selasphorus platycercus, and Archilochus alexandri hummingbirds.
Variables
- Bird ID: Banding code and Id#
- Capture date: calendar date
- Day of Year (DOY): sequential date
- Capture time: clock time AM/PM
- Age: AHY = after hatch-year (adult), HY = hatch-year (juvenile)
- Sex: M = male, F = female
- δ2Hf: feather hydrogen stable isotope ratio (‰)
- δ2Hcor: discrimination factor-corrected hydrogen stable isotope ratio (‰)
- Mb (g): body mass (grams)
- Mf (mg): fat mass (milligrams)
- Mf (%): fat mass as percent of body mass
- Wing Chord (mm): direct distance from leading edge to trailing edge of wing (millimeters)
- Fat-free mass (g): Mb - (Mf / 1000) (grams)
- Log Torpor: 0 = no torpor, 1 = torpor, n/a = not applicable
Note: cells marked with "n/a" represent birds that were not measured in respirometry trials.
File: Supplemental_Material_S2_Summary_of_Body_Mass_Components.csv
Description:
Tabular summary of body mass components in the four species of hummingbirds we examined.
Variables
- Species
- Age - Sex (M, F = adult, m, f = juvenile)
- Sample size (N)
- Average body mass in grams (Mb)
- Average fat-free mass in grams (Mff)
- Average Body fat in grams (Mf)
- Average body fat as percent of Mb
Table Caption
This table summarizes diverse patterns in mean Body Mass (Mb), Fat-free Mass (Mff), Fat Mass (Mf), and % Fat by age and sex, in the four study species. Adult S. rufus hummingbirds had higher Mb than juveniles (F(2, 376)=27.3, p<0.0001), and female S. rufus had higher Mb than males (F(2, 376)=5.5, p=0.0199). Adult S. calliope had higher Mb than juveniles (F(2, 67)=6.9, p=0.0109), and female S. calliope had higher Mb than males (F(2, 67)=4.2, p=0.0444). In S. platycercus, adults had higher Mb than juveniles (F(2, 202)=5.4, p=0.0212), and there was no sex difference (p=0.1640). In A. alexandri, Mb did not differ with age (p=0.8771), and females had higher Mb than males (F(285)=24.9, p<0.0001. S. rufus adults had higher Mff than juveniles (F(2,376)=42.7, p<0.0001), and there was no sex difference (p=0.1507). Mff did not differ with age or sex in S. calliope or S. platycercus (p>0.05 in all variables). In A. alexandri, there was an interaction with age and sex, where females had higher Mff than males (F(3, 284)=73.9, p<0.0001), and adult females had higher Mff than any other age/sex class (F(3, 284)=10.1, p=0.0017). In both S. rufus and S. platycercus, Mf was higher in adults (F(2, 376)=42.7, p<0.0001 and F(2, 202)=14.3, p=0.0002) respectively, with no effects of sex (p>0.05). No effects of age or sex were observed in S. calliope or A. alexandri (p>0.05). Results of % Fat versus age and sex mirrored those of Mf in grams. Upper case (M, F) represents adult males and females, and lower case (m, f) represents juvenile males and females. Values are averages ± standard error.
File: Supplemental_Figure_S3_Adult_d2H_vs_Date_and_Sex.pdf
Description:
Corrected feather hydrogen stable isotope ratios (δ2Hcor) in adult hummingbirds did not change with capture Day of Year (DOY). Male S. rufus (A) had significantly depleted feathers compared to females (-4.0 ± 1.5 ‰). Blue and red lines (A) represent male and female values in S. rufus, and purple lines in (B), (C), and (D) are the output of model averaging, where the effect size of sex was too small to detect graphically. Violin plots depict distributions of δ2Hcor in each species.
Code/software
No free or open software was used in this study.
Access information
Other publicly accessible locations of the data:
- None
Data was derived from the following sources:
- Data collection by the author in the field
We captured 942 adult and juvenile hummingbirds of four species at a stopover site near Mimbres, NM, USA, between July and September 2022. A Quantitative Magnetic Resonance scanner was used to non-invasively measure lipid content in each bird. A Push Flow-through Respirometry System was used to measure overnight metabolic parameters and use of nocturnal torpor in 94 adult and juvenile birds. Respirometry metabolic data were recorded using Sable Systems ExpeData version 1.9.27 software. A retrix (R1) feather was collected from each bird, and feather hydrogen stable isotope ratios were measured on 942 specimens, at the University of New Mexico Center for Stable Isotopes. Results for each species were compiled into analyses of capture date, body mass, fat-free mass, fat mass, molt/natal origin, torpor probability, and torpor fat thresholds. Results of a subset of 63 juvenile birds tested in respirometry were used to predict torpor probabilities in the larger dataset of 573 juvenile birds, in which we measured both lipids and feather hydrogen stable isotopes, but not metabolic parameters. All birds were fed and released at the capture site following measurements. Microsoft Excel and JMP Pro 18 Statistical Software were used to perform data analyses. We conducted statistical analyses to determine how lipid-loads and torpor patterns are related to natal origins and migration patterns.